Advances in digital technology have found increasingly wide application in the field of medicine in general and radiology in particular. A two-dimensional radiographic apparatus for radiological use has been developed in order to digitalize X-ray images, in which a scintillator is used to convert the X-rays into visible light that is then sensed and formed into a diagnostic image by image sensors.
As two-dimensional radiographic apparatuses, compact CCD image sensors for use in dentistry have already been commercialized, and for mammography and thoracic X-ray use a large-scale, still-image sensing apparatus using maximum 43 cm-square panels of amorphous silicon has been produced. Image sensors that use amorphous silicon semiconductors formed on a glass substrate can be formed easily into large panels, and large-scale radiographic apparatuses have been achieved using four tiles of such panels. An example of this type of technology is described in U.S. Pat. No. 5,315,101.
Similarly, a large-panel radiographic apparatus comprising a plurality of monocrystal image sensors (such as silicon image sensors) has been proposed. An example of this type of technology is U.S. Pat. No. 5,159,455 shown in FIG. 12. Silicon-based CCD-type image sensors and MOS- or CMOS-type image sensors may be used for the monocrystal image sensors.
Further advances in the digitalization of medical radiographic diagnostics are expected, in the form of still more sensitive still-image sensing apparatuses and next-generation, moving-image sensing apparatuses.
In this case, acquiring a moving image necessitates directing a continuous X-ray onto a human subject. The known adverse effects on living tissue of prolonged exposure to X-ray radiation, however, make it desirable to reduce the intensity of X-ray to, e.g., 1/100 of normal intensity and to employ read speeds of 60-90 frames/sec, which in turn requires apparatuses that are several tens of times faster and more sensitive than ordinary still-image acquisition equipment.
The process of manufacturing an amorphous silicon panel image-sensing apparatus possesses the advantage of yielding larger panels compared to the process for manufacturing CCD image sensors and CMOS image sensors, but with the disadvantage that it is more difficult to carry out fine processing of a semiconductor on a glass substrate than on a monocrystal silicon semiconductor substrate, and as a result the output signal line capacity increases. This capacity is the largest source of noise (kTC noise) and limits improvements in sensitivity. Moreover, with amorphous silicon the semiconductor characteristics are not enough to increase the speed of operation, so that acquisition of moving images at speeds of 30 frames/sec or more is difficult.
CCD image sensors, though of the complete-depletion type and therefore sensitive, are unsuited as wide-area image-sensing elements. A CCD image sensor is a charge transfer device, so as the area (i.e., the number of pixels) increases and the number of transfer steps grows large, transfer becomes a problem. In other words, the drive voltage is different at the drive terminal and near the center, making complete transfer difficult. In addition, power consumption, which may be expressed as CVf2 (where C is the capacity across the substrate and the well, V is the pulse amplitude, and f is the pulse frequency) experiences a ten-fold increase compared to that of a CMOS image sensor because C and V increase as the area increases. As a result, the drive circuitry in this area generates heat and noise, degrading the S/N ratio. For these reasons a CCD-type image sensor is not suitable as a large-scale image sensor.
In a simple large-panel image-sensing apparatus using a multiplicity of monocrystal image sensors, dead space is inevitably created wherever the image sensors adjoin (necessitated by the need for a region separate from the pixel region for providing peripheral circuitry such as a shift register, multiplexer and amplifier, external terminals for transmitting signals and power to and from an external device and a protective circuit composed of a protective diode or a protective resistance against static electricity). This dead-space portion leads to line defects (that is, gaps in the image) and a deterioration in picture quality. For this reason a tapered FOP (fiber optic plate) is used to direct light from a scintillator around the dead spaces and toward the image sensor pixel region. However, such a configuration requires a very expensive tapered FOP, which increases production costs. Moreover, a tapered FOP has a disadvantage in that the sharper the angle of taper the harder it is for light from the scintillator to enter the tapered FOP, which leads to a decrease in output light level that can offset the sensitivity of the image sensors and reduce the overall sensitivity of the apparatus.